3 RISK PREDICTION ASSESSMENT AND distribute...Chapter 3 • Assessment and Risk Prediction 53conviction and sentencing, the probation officer will complete the PSI, which will include

52

ASSESSMENT AND RISK PREDICTION3

LEARNING OBJECTIVES

1. Understand the importance of the presentence investigation report (PSI) and why accurate information is critical to later supervision and treatment needs of the offender.

2. Understand basic concepts inherent to assessment.

3. Know the meaning of false positives, false negatives, true positives, and true negatives.

4. Understand the difference between static and dynamic risk factors.

5. Identify subjective and objective means of assessment and risk prediction.

6. Be familiar with the Wisconsin Risk Assessment and the flaws associated with that instrument.

7. Be familiar with the Level of Service Inventory–Revised, and understand the strengths associated with that instrument.

8. Understand the reasons why the MMPI-2 Criminal Justice and Correctional Report is presented as a premier instrument.

INTRODUCTION

Because of concerns with public safety, it is imperative that a correc-tional agency be as adept as possible at accurately assessing the future prognosis of any offender under its supervision. For the purposes of this text, the term prognosis refers to the likelihood that an offender will successfully reform and will refrain from further criminal activity. Thus, there is both a treatment component (reform) and a public safety component (likelihood of further criminal activity) contained within an offender’s prognosis. Both must be considered simultaneously if the correctional agency is to fulfill its mission in a satisfactory manner.

The intake is the process that occurs when the offender is ini-tially entered into the correctional system. This process is heavily tied to information obtained from the presentence investigation report (PSI) that the probation department will provide to the presiding judge of an offender’s case. The presentence investigation report is the file that includes a wide range of background information on the offender. This file will typically include demographic, vocational, educational, and personal information on the offender as well as records on prior offending patterns and the probation department’s recommendation as to the appropriate type of sentencing and super-vision for the offender in question. In many respects, the PSI is the initial point of assessment, and it will often be utilized during assess-ment in the institutional setting or when the offender is officially placed under the jurisdiction of the probation department. Because the PSI is used in all community supervision agencies and because this is a centerpiece component of the probation process, we will now take a closer look at the PSI in the section that follows.

THE PRESENTENCE INVESTIGATION REPORT (PSI)

The primary purpose of the presentence investigation report is to provide the court with the necessary information from which a sentencing decision can be derived. The PSI is conducted after a defendant is found guilty of a charge (whether by pleading or court finding) but prior to the point of sentencing. During this point betwixt

Copyright ©2019 by SAGE Publications, Inc. This work may not be reproduced or distributed in any form or by any means without express written permission of the publisher.

Do not

copy

, pos

t, or d

istrib

ute

Chapter 3 • Assessment and Risk Prediction 53

conviction and sentencing, the probation officer will complete the PSI, which will include extensive information pertaining to the offender. This information, along with a sentencing recommendation, will aid the judge who must ultimately fashion a sentence as well as any corollary obligations attached to that sentence.

Likewise, the PSI tends to serve as a basic foundation for supervision and treatment plan-ning throughout the duration of the offender’s sentence. This document will serve as a refer-ence point for placing the offender in a variety of programs; it is used when the offender is on supervision and may serve as a guidepost for jail and detention facilities as well. The PSI will contain sundry amounts of information about an offender’s background, including education, social and medical history, and work experience. In addition, an offender’s file may contain a number of documents from a variety of professionals who provide services for the community supervision agency. Community supervision officers should make a point to keep the informa-tion as confidential as is realistically possible. Thus, free and open disclosure of the informa-tion should be discouraged.

The PSI is typically conducted through an interview with the offender. Because the PSI information is largely obtained from the interview process, it is naturally important that com-munity supervision officers have good interviewing skills. This cannot be overstated given the fact that community supervision officers are in contact with persons on a routine basis from whom they must collect and record information. Further, the community supervision officer will often conduct interviews with family members, employers, and so forth to validate the information received from the offender. It is of course important that the PSI contain reliable information that is relevant to the defendant’s character, attitude, and activities.

According to the American Probation and Parole Association (APPA, 2006), the PSI should reflect the background of the offender and should be as accurate and comprehensive as the probation department’s resources will allow. In being comprehensive, a report should contain certain bits of information that are germane to effective sentencing and offender placement. According to the APPA, the specific items included in most reports are as follows:

1. A complete description of the offense and circumstances surrounding it, not limited to the aspects developed for the record as part of the determination of guilt

2. A statement from the victim and a description of the victim’s status, the impact upon the victim, losses suffered by the victim, and restitution due the victim

3. A full description of any prior criminal record of the offender

4. A description of the educational background of the offender

5. A description of the employment background of the offender, including military record and his or her present employment status, financial status, and capabilities

6. The social history of the offender, including family relationships, marital status, interests and activities, residence history, and religious affiliations

7. The offender’s medical history and, if desirable, a psychological or psychiatric report

8. Information about environments to which the offender might return or to which he or she could be sent should probation be granted

9. Supplementary reports from clinics, institutions, and other social agencies with which the offender has been involved


Do not

copy

, pos

t, or d

istrib

ute

54 Essentials of Community Corrections

10. Information about special resources that might be available to assist the offender, such as treatment centers, residential facilities, vocational training services, special educational facilities, rehabilitation programs of various institutions in which the offender might be committed, special programs in the probation department, and other similar programs that are particularly relevant to the offender’s situation

11. A summary and analysis of the most significant aspects of the report, including specific recommendations as to the sentence (A special effort should be made in the preparation of presentence reports not to burden the court with irrelevant and unconnected details.)

Figure 3.1 provides a good illustration of a typical PSI that may be submitted to a court prior to sentencing. When examining the PSI form from the state of Kansas, it is clear that most of the information recommended by the APPA is included. The form in Figure 3.1 is very well structured and largely provides this information in a standardized format. The typical PSI will also tend to have narrative components that are provided by the offender as well as others who are interviewed by the presentence investigator.

Many states have rules of confidentiality when gathering privileged information for the PSI as well as regulations regarding the disclosure of that information once it has been included in the final report. However, the conversations between the probation officer and the defendant are not considered privileged, and such information may be subject to disclosure. In addition, when the conviction information is a matter of public record, probation officers may disclose this information to noncriminal justice agencies or other persons as long as this disclosure does not violate any agency regulations. As one might reasonably suspect, while

in the course of their duty, probation officers may disclose an offender’s information to criminal justice agencies on a need-to-know basis in most states. Courts will typically base their decision of disclosure on the type of crime that was committed as well as the probation officer’s assessment of the offender’s potential to harm other victims. Finally, most states do allow the probation officer to provide information or records to the offender’s victims, but often these same records are not avail-able to the general public.

In the 1949 case of Williams v. New York (337 U.S. 241), the U.S. Supreme Court upheld the confidentiality of the PSI. This is largely due to the perception of the Court that the pre-sentence investigator is a neutral and detached party with no real vested interest in the punishment of the offender. Indeed,

the probation officer was considered to be a helping professional rather than one in league with prosecutorial efforts. Nevertheless, many states do allow defendants to view their PSI and have even passed regulations requiring that defendants be given the chance to review their PSI and to refute its contents before the sentence phase begins.

In the federal probation system, the contents of the PSI are disclosed to the defendant, the defendant’s legal counsel, and the prosecuting attorney. Likewise, since the passage of the 1984 Sentencing Reform Act, the content of the PSI has determined the parameters of sentencing. This means that the central nature of the PSI to the sentencing process has become a source of scrutiny by defendants since it is important to the future sentence and since it serves as the basis for participation in specific programs offered within a prison system or community

PHOTO 3.1 Probation and parole officer Chris Byrd and a court attorney discuss some of the content in an offender’s PSI before the sentencing judge appears in the courtroom. In some cases, negotiations may be arranged by various court actors depending on the circumstances of the crime and the offender.


Do not

copy

, pos

t, or d

istrib

ute


FIGURE 3.1 ■ Presentence Investigation Report in the State of Kansas

2004 KANSAS SENTENCING GUIDELINES—PRESENTENCE INVESTIGATION REPORT FACE SHEET 7.04

Judicial District: ____________________________________

County: ___________________________________________

Case No: _________________________________________

Name: _____________________________________________

A/K/A’s: ___________________________________________

D.O.B.: ____/____/____ S.S.N. #: _________________

Age: _______ K.B.L. No: _________________________

Sex: Male Female Race: W B A.I. A

Ethnicity: Hispanic Non-Hispanic

Address: ____________________________________________________________________________________________________________________________________

Citizenship: U.S. Citizen of: ____________

DETAINER OR OTHER CHARGES PENDING? YES NO

SUBJECT IN CUSTODY AWAITING SENTENCING? YES NO

begin _____ end _____

begin _____ end _____

IF OFFENDER WAS UNDER 18 YEARS OF AGE WHEN CRIME(S) WAS COMMITTED AND WAS TRIED AS AN ADULT, OFFENDER WAS:

ADJUDICATED AS AN ADULT UNDER KSA 38-I636 AUTOMATICALLY CONSIDERED AN ADULT

BECAUSE OF A PRIOR FELONY

NAMES OF CO-DEFENDANTS, IF ANY _________________________________________________________

NAME OF DEFENSE ATTORNEY: _______________________________________________________________

TYPE OF COUNSEL PRIOR TO SENTENCING: RETAINED APPOINTED SELF OTHER

NAME OF PROSECUTING ATTORNEY: __________________________________________________________

NAME OF SENTENCING JUDGE: _______________________________________________________________

DATE OF GUILTY PLEA OR JUDGMENT: ____/____/____

DATE OF SENTENCING: ____/____/____

NAME OF PRESENTENCE INVESTIGATOR:_______________________________________________

DATE ASSIGNED ____/____/____

DATE SUBMITTED: ____/____/____

OFFENSES

NAME OF PRIMARY OFFENSE: __________________________________________K.S.A. No:_____________ Felony Offgrid Misd. Nongrid

Person Drug Attempt Conspiracy

Nonperson Nondrug Solicitation

Severity Level: ___ Criminal History Score: ___

Sentencing Range:

Standard _____ Presumptive Prison

Aggravated _____ Presumptive Probation

Mitigated _____ Border Box

Special Rule Applicable (see p. 4)

Mandatory Drug Treatment (“SB 123”) Drug Treatment with Court finding Not eligible for Drug Treatment due to criminal history

Postrelease Supervision Duration: 12 months 24 months 36 months 60 months No Postrelease - per K.S.A. 22-3716 (e)

Probation Duration: 12 months 18 months 24 months 36 months Other

NAME OF ADDITIONAL OFFENSE:_______________________________________________

K.S A. No: ________________ Felony Offgrid Misd. Nongrid

Person Drug Attempt Conspiracy Nonperson Nondrug Solicitation

Severity Level: ____ Criminal History Score: ____

Sentencing Range:

Standard ____ Presumptive Prison

Aggravated ____ Presumptive Probation

Mitigated ____ Border Box


Mandatory Drug Treatment (“SB 123”) Drug Treatment with Court finding Not eligible for Drug Treatment due to criminal history


Probation Duration: 12 months 18 months 24 months 36 months Other _________

KSG Desk Reference Manual 2004 Appendix D Page 1

(Continued)


Do not

copy

, pos

t, or d

istrib

ute


2004 KANSAS SENTENCING GUIDELINES—PRESENTENCE INVESTIGATION REPORT FACE SHEET SUPPLEMENTAL PAGE

NAME OF ADDITIONAL OFFENSE:

K.S.A. No: _____________ Felony Offgrid Misd. Nongrid



Severity Level: ____________ Criminal History Score: ___

Sentencing Range:

Standard _____ Presumptive Prison

Aggravated _____ Presumptive Probation

Mitigated _____ Border Box


Mandatory Drug Treatment (“SB 123”)

Drug Treatment with Court finding

Not eligible for Drug Treatment due to criminal history



NAME OF ADDITIONAL OFFENSE:

K.S.A. No: _____________ Felony Nongrid

Misd. Offgrid



Severity Level: _______________ Criminal History Score: ___

Sentencing Range:

Standard _____ Presumptive PrisonAggravated _____ Presumptive ProbationMitigated _____ Border Box








K.S A. No: ________________ Felony Nongrid

Misd. Offgrid




Sentencing Range:








Postrelease Supervision Duration: 12 months 24 months

36 months 60 months No Postrelease - per K.S.A. 22-3716 (e)

Probation Duration: 12 months 18 months

24 months 36 months Other _________


K.S A. No: ________________ Felony Nongrid

Misd. Offgrid




Sentencing Range:








Postrelease Supervision Duration: 12 months 24 months

36 months 60 months No Postrelease - per K.S.A. 22-3716 (e)

Probation Duration: 12 months 18 months

24 months 36 months Other _________


FIGURE 3.1 ■ (Continued)


Do not

copy

, pos

t, or d

istrib

ute


KANSAS SENTENCING GUIDELINES—PRESENTENCE INVESTIGATION REPORT CURRENT OFFENSE INFORMATION

OFFICIAL VERSION:

DEFENDANT’S VERSION:

VICTIM’S INJURY / DAMAGE / STATEMENT(S):

TOTAL RESTITUTION ___________________________________________________________________________

OWED TO Name: ___________________________________________________________________________

Address: ___________________________________________________________________________

___________________________________________________________________________

Amount: ___________________________________________________________________________

Name: ___________________________________________________________________________

Address: ___________________________________________________________________________

___________________________________________________________________________

Amount: ___________________________________________________________________________

Name: ___________________________________________________________________________

Address: ___________________________________________________________________________

___________________________________________________________________________

Amount: ___________________________________________________________________________

Name: ___________________________________________________________________________

Address: ___________________________________________________________________________

___________________________________________________________________________

Amount: ___________________________________________________________________________

Name: ___________________________________________________________________________

Address: ___________________________________________________________________________

___________________________________________________________________________

Amount: ___________________________________________________________________________

STATEMENT(S):


(Continued)


Do not

copy

, pos

t, or d

istrib

ute



RECOMMENDED PLACEMENT: [The following is not a recommendation regarding the appropriate disposition to be imposed in this case, but is provided as the court services or community corrections officer’s professional assessment if the court places the offender on probation or orders some form of community sanction.]

______ COMMUNITY CORRECTIONS [Indicate the criteria from K.S.A. 75-5291(a)(2) met by this defendant qualifying him/her for placement in Community Corrections.]

______ COURT SERVICES Mandatory Drug Treatment (“SB 123”) (a)(2)(A) Listed grid box (Border box, level 6 H or I, level 7 C-I)______ DEPARTMENT OF CORRECTIONS (a)(2)(B) Downward dispositional departure (presumptive prison) (a)(2)(C) Severity level 7 or higher sex offender (a)(2)(D) Condition violator (a)(2)(E) Scored “high risk or needs, or both” (a)(2)(F) Follows successful completion of a conservation camp program

OFFICER’S ASSESSMENT OF CONDITIONS OF PROBATION:

K.S.A. 21-4610 and; [Check All That Apply] (Alcohol) (Drug) (Mental Health) evaluation (follow recommendations) (In) (Out) Patient (Alcohol) (Drug) (Mental Health) treatment (Follow recommendations of counselor) (AA) (NA) Attendance No possession or consumption of alcohol or illegal drugs Submit to random (Breath) (Blood) (Urinalysis) testing at request of C.S.O. at defendant’s own expense Community Service Work [ _____ Hours] (Gain) (Maintain) employment Notify the C.S.O. of changes in employment, residence and phone number No contact with (victim) (co-defendant) Educational program - (G.E.D.) (Vocational) (Higher Education) Curfew Restriction: __________________________________________________________________ Travel Restriction: ___________________________________________________________________ OTHER ____________________________________________________________________________ OTHER ____________________________________________________________________________ OTHER ____________________________________________________________________________ OTHER ____________________________________________________________________________ OTHER ____________________________________________________________________________

PLEASE CHECK ANY SPECIAL SENTENCING RULES APPLICABLE TO THIS CASE:

Person Felony Committed with a Firearm – K.S.A 21-4704 (h) (Shall be presumed imprisonment) Aggravated Battery on an L.E.O. – K.S.A. 21-4704 (g) (Shall be presumed imprisonment) (6-H or 6-I) Aggravated Assault on an L.E.O. – K.S.A. 21-4704 (g) (Shall be presumed imprisonment) (6-H or 6-I) Crime Committed for Benefit of Criminal Street Gang – K.S.A. 21-1704 (k) (Shall be presumed imprisonment) Persistent Sex Offender – K.S.A. 21-4704 (j) (Shall be presumed imprisonment) Felony DUI (third) – K.S.A 21-4704 (i) Felony DUI (fourth or subsequent) – K.S.A. S-1567 (g) Felony Criminal Deprivation of Property / Motor Vehicle [Crime committed prior to July 1, 1999.] – K.S.A. 21-3705 (b) Felony Domestic Battery – K.S.A. 21-3412a (b)(3) Crime Committed While Incarcerated and Serving a Felony Sentence, or While on Probation, Parole, Conditional Release or Postrelease Supervision for a Felony – K.S.A. 21-4603d (f) (New sentence shall be consecutive) – K.S.A. 21-4608 Crime Committed While on Felony Bond – K.S.A. 21-4603d (f) (Crime committed on or after 7/1/99 may sentence to prison even if presumptive probation) Kansas Securities Act – K.S.A. 17-1267 Extended Jurisdiction Juvenile Imposed – K.S.A. 38-1636 Second or Subsequent Manufacture of a Controlled Substance Conviction – K.S.A. 21-4705 (e) Residential Burglary After a Prior Residential or Nonresidential or Agg. Burglary Conviction – K.S.A. 21-4704 (1) (Shall be presumed imprisonment) Second Forgery – K.S.A. 21-3710 (b)(3) Third or Subsequent Forgery – K.S.A. 21-3710 (b)(4) Mandatory Drug Treatment – K.S.A. 21-4729 (SB 123) Other ____________________________________


Source: Kansas Sentencing Commission. (2004). Kansas sentencing guidelines: Presentence investigation report. Topeka, KS: Author. Retrieved from http://www.kspace.org/bitstream/1984/69/7/Appendix_D_2004_PSI_Form.pdf.


Do not

copy

, pos

t, or d

istrib

ute


supervision agency. In many cases, the PSI serves as the basis of classification within both institutional and community-based systems, and this is especially true in the federal system.

Thus, the PSI is a significant document for the offender, and its content must be closely guarded to ensure accuracy. From an assessment standpoint, this is an important consideration because this critical document is completed by one person (the presentence investigator) and thus allows for a great deal of discretion by that person. Given that this document will follow the offender throughout his or her sentence, it is perhaps one of the most important components of the community supervision process. The accuracy of information is critical to this stage of the community supervision process, and the information contained must pass the scrutiny of both the prosecution and the defense. Otherwise, this means of assessment/classification is greatly compromised. This point is illustrated in Section 32.1, subsections b through f, of the U.S. Courts Sentencing Guidelines (U.S. Sentencing Commission, 2016), which read as follows:

Presentence Investigation Report. The probation office shall prepare a presentence investigation report in every case unless the court finds that sufficient information exists in the record to enable the meaningful exercise of its sentencing authority pursuant to 18 U.S.C. § 3553. The probation office, during the presentence investigation, shall provide notice and a reasonable opportunity to defendant’s counsel to attend any interview of the defendant.

Written Version of Facts. No later than fourteen (14) days following a plea or verdict of guilty, the government shall provide the probation office with a written version of the facts of the case, including all relevant conduct. The government shall provide, at a minimum, the probation office with the same discovery materials it provided to the defendant. The prosecutor assigned to the case and the primary case agent shall make themselves reasonably available to the probation office to answer any inquiries.

Disclosure of Presentence Investigation Report. No later than thirty-five (35) days prior to the scheduled sentencing date, the probation officer shall disclose the initial presentence investigation report to the parties. One copy shall be given to counsel for the government. Two copies shall be given to defense counsel, who shall give one copy to the defendant for review. Defense counsel shall ensure that the defendant has timely reviewed and understands the presentence report.

Objections to Presentence Investigation Report. No later than fourteen (14) days after receiving the initial presentence report, counsel for the government and counsel for the defendant shall deliver to the probation officer, and to each other, written objections of fact or guideline application to the initial presentence report. If counsel has no objections, counsel shall so notify the probation officer in writing. Delivery of said objections shall be made by mail, in person, or by facsimile transmission. A party waives any objection to the presentence report by failing to comply with this rule unless the court determines that the basis for the objection was not reasonably available prior to the deadline.

Revised Presentence Investigation Report and Addendum. If either party objects to the presentence report, the probation officer shall conduct such further inquiry as is necessary to attempt to resolve the objections raised.

This is an important point of discussion in relation to the PSI. Aside from the fact that the PSI will contain various information that is related to the supervision and treatment of the


Do not

copy

, pos

t, or d

istrib

ute


offender, the accuracy of the information is sometimes questionable and can actually impair the treatment outcomes of offenders. This is the primary concern in noting the use of the PSI. As will be seen in a later section of this chapter, other departments (e.g., in Travis County, Texas) have noted inefficiency in their presentence investigation process and have had to streamline and revamp their systems. Likewise, it will be seen in Chapter 4 that the completion of the PSI is a large part of the community supervision officer’s job design, and the excessive paperwork associ-ated with the PSI and other documentation serves as a primary source of stress for community supervision officers. Being that judges tend to rely on the recommendation of presentence inves-tigators when determining offenders’ sentences, it is clear that the PSI is a critical aspect of the community supervision process from both a reintegration and a public safety perspective. More will be discussed regarding the PSI process later in this and other chapters. However, students should consider the following two example scenarios, presented by Michael Santos (2000):

Though these are just two examples, many others abound. Further, such instances are even more likely to occur within state and county systems where community supervision caseloads are even higher and where turnover among staff is higher as well. Aside from understanding the basic information that is contained in a PSI, the fact that community supervision officers collect the information in the PSI, and the basic use of that information, the single most important factor for students to understand in regard to the PSI is that it is only as good as the investigator allows it to be.

EXAMPLE 1: ALFREDO

Alfredo, an offender who was convicted of conspiracy to distribute cocaine, played a minor role in his offense. He allowed others to use his telephone to facilitate their drug transactions. Alfredo was not privy to the quantity of drugs being sold, nor [was he privy] to the number of transactions that took place over his telephone line. Yet the PSI report indicated over 20 kilograms of cocaine were sold and that all conspirators, including Alfredo, were equally culpable.

The sentencing judge, who had listened to all the testi-mony at trial, however, knew that Alfredo was a minor player in the conspiracy. He found Alfredo less culpable than the oth-ers and gave him a downward departure from the sentencing guidelines because of his minor role. The PSI, however, was never amended. As a result, in using the PSI as its reference point, the BOP [U.S. Bureau of Prisons] classifies Alfredo as a serious offender and refuses him camp placement. Alfredo has tried to have his PSI amended several times during his confinement, but the judge has ruled the matter moot because Alfredo received the downward departure at sentencing.

This was Alfredo’s first experience in the criminal justice system. He did not appreciate the significance of his PSI at sentencing and relied on his public defender to represent him in all matters. The public defender was successful in persuading the judge that the PSI inaccu-rately portrayed Alfredo as an equal participant in the conspiracy, but made no attempt to change the PSI itself. Accordingly, Alfredo was sentenced appropriately, but he serves his sentence in more severe conditions than other similarly situated offenders because his PSI remains inaccurate.

Alfredo has made efforts to show his case manager the sentencing transcripts where the judge clearly ruled that Alfredo was less culpable than the others and sentenced him accordingly. Such evidence is irrelevant to the BOP, however, as the PSI governs all classification decisions. Consequently, Alfredo serves his sentence in a higher-security facility and he may be denied access to half-way house placement toward the conclusion of his term.

EXAMPLE 2: RICH

Rich pled guilty to an indictment charging several defen-dants with organized crime involving extortion and murder. Rich’s role, however, was minor and he was sentenced to serve approximately five years as a result of his convic-tion; other codefendants who were charged on Rich’s same indictment received sentences of life imprisonment.

Rich’s probation officer conducted the presentence investigation for all defendants on the indictment. When Rich appeared for his interview, he was accompanied by his defense counsel. The defense counsel heard every-thing said during Rich’s interview, and Rich was cooperative throughout the proceeding. When the report was completed


Do not

copy

, pos

t, or d

istrib

ute


THE BASICS OF RISK ASSESSMENT

Before we can begin a discussion on specific instruments and inventories, it is important that the student understand the underlying presumptions behind risk prediction. In any type of risk assessment, there are some common principles that tend to run consistently through-out. If these basic statistical considerations or methodological processes are not honored during the construction of an instrument, the instrument is likely to be flawed and will there-fore be less accurate in prediction. This places the public at risk and also improperly classifies offenders, even when they are not in a position to be dangerous to the public. James Austin (2006) provides six basic suggestions for correctional officials who wish to know whether their instruments are effective. Most of Austin’s comments center on the methodology that is utilized in constructing and testing the instrument, thereby relating to the validity and reliability of that given instrument. Austin provides a very good, hard-hitting description. Further, it should be pointed out that all of the recommendations by Austin are consistent with the notion of evidence-based practices, which will be discussed in future chapters, especially Chapters 4 and 13. Austin’s recommendations are as follows:

1. Risk assessment instruments must be tested on your correctional population and separately normed for males and females.

Austin (2006) notes that when assessment tools are tested on the offender populations in one area of the nation, they may not be as relevant to offenders in another area. For example, take the state of Florida as compared with the state of Iowa. Is it likely that the correctional populations are similar? These issues should be considered, and agencies should use instruments that are essentially normed for, or tailored to, the characteristics of their own offender populations. As Austin notes, “in research terms this issue has to do with the ‘external validity’ of the instrument and the ability to generalize the findings of a single study of the instrument to other jurisdictions” (p. 1). Thus, if an instrument is normed for an offender population that is different from the one being examined, it is likely that the assessment will not be accurate.

Further, male and female offenders are not the same. The issues leading to the criminality, types of crimes committed, and prognosis for treatment of female offenders all tend to be different from those of male offenders. Thus, risk assessments for both types of offenders should also be different, yet in many cases they are not. At the very least, assessment tools must give appropriate weight to gender differences among offenders. Austin (2006) notes that “recidivism and career criminal studies consistently

and given to Rich and his attorney for review, however, it was clear that the probation officer had confused some of Rich’s codefendants’ recalcitrant statements and inappro-priately applied them to him. The PSI indicated that Rich was involved in murders, domestic abuse, and drug sales. In fact, none of this information applied to Rich.

At the sentencing hearing, Rich’s attorney succeeded in showing the clear error in the PSI. He ordered the PSI to be amended, and it was changed. Rich was sentenced appropriately.

When Rich reported to prison, however, he learned that his case manager was using the original, erroneously

prepared PSI. Consequently, the case manager told Rich that he would never be eligible for camp placement and that he may not be eligible for half-[way] house placement either.

Rich contacted his attorney, who then initiated legal action to force the BOP to use the corrected ver-sion of the PSI. The court has thus far refused to grant the order, though, stating that such issues should have been resolved at the sentencing hearing. And as a result of Rich’s attorney’s error, he continues to serve his sen-tence under the misclassification wrought by an inaccu-rate PSI.


Do not

copy

, pos

t, or d

istrib

ute


show that females are less involved in criminal behavior, are less likely to commit violent crimes, and are less likely to recidivate after being placed on probation or parole” (p. 1).

2. An inter-rater reliability test must be conducted.Austin (2006) contends that both an inter-rater reliability test and a validity test

must be completed by independent researchers who have no monetary or political incentive in regard to the testing outcome. Though these terms may be alien to many undergraduate students of criminal justice (particularly if a student has not had a class in research methods), they are important to understand. As applied to our current discussion, inter-rater reliability has to do with the consistency of the results that are obtained from an instrument. It should consistently yield the same outcomes regardless of the person who is using the instrument, presuming that each person administering the assessment is competent in administering it.

3. A validity test must be conducted.Likewise, instruments must be valid. Validity simply ensures that the instrument

is actually measuring what researchers believe it is measuring. In some cases, instruments can provide measures that correlate with a given issue, but the cause of that correlation may be due to some unknown factor. With respect to our current discussion, it is important that instruments actually measure recidivism (or perhaps reintegration, depending on our intent) rather than some other unknown variable. Consider that states’ economic conditions may be stable or changing. Good economic conditions are associated with less offending. If an instrument in this area predicts less likelihood of recidivism among that area’s offender population, how do we know whether this is due to actual characteristics of the offender or to transient economic conditions that could change a year after the offender is released? This issue must be addressed appropriately for accurate assessment.

4. The instruments must allow for dynamic and static factors that have been well accepted and tested in a number of jurisdictions.

The use of dynamic and static risk factors will be discussed in depth in a subsection that follows. Examples of dynamic risk factors include characteristics such as age, marital status, and custody level. These characteristics can and do change over time. Static risk factors include characteristics such as age at first arrest, crime seriousness, and prior convictions. Once established, these characteristics do not fluctuate over time.

5. The instruments must be compatible with the skill level of your staff.As noted in the second suggestion above, the accuracy of the assessment is dependent

on the skill of the person administering the assessment. Often, agencies use simple risk-assessment tools because these are easy for their staff to understand and administer with little need for extensive training. While on paper this may seem to be effective, and while it may appear that agency staff are using objective criteria effectively, they are often not trained to utilize more sophisticated instruments that have better levels of reliability and validity. Thus, instruments tend to be poor or, if the agency does utilize well-developed instruments, the staff may be poorly trained in their administration. Thus, even effective instruments can produce ineffective or inaccurate results. In addition, staff must have experience administering these instruments. Training alone is not sufficient, as there is no replacement for the self-assurance that comes with repetition.

6. The risk assessment must have “face validity” and transparency with staff, prisoners, probationers, parolees, and policy makers.


Do not

copy

, pos

t, or d

istrib

ute


The instrument and the process of assessment must be understood and recognized as credible by all persons in the agency. Instruments that are only understood by “eggheads” and academics will never go over well with most practitioners. Further, if the instrument is perceived as being too “bookish” in nature and not applicable to the realities of the “street,” so to speak, both practitioners and offenders will see the instrument as artificial and sterile, and not really able to probe the true reality of what an offender may (or may not) do. This can also be exacerbated by the presence of false positives (falsely predicting recidivism when it would not have occurred) or false negatives (failing to predict recidivism when it does later occur). The student should understand that a lack of “face validity” means that the instrument is not recognized as valid on its face, or at initial glance, by those who judge its ability to truly discern a set of characteristics. In this case, if the instrument appears to ask bogus questions or if the questions are asked in a naive manner, it will not be perceived as credible by practitioners and offenders alike.

FALSE POSITIVES AND FALSE NEGATIVES

Agency administrators tend to assume that their instruments are, in fact, valid and reliable and that their staff are able to effectively use these instruments. To some extent, these admin-istrators are forced to operate under these assumptions, as alternatives may be unavailable or unacceptable. Further, these administrators operate under these assumptions when mak-ing release decisions. However, as has just been discussed, when instruments are not normed appropriately, do not possess appropriate degrees of external validity, or lack other forms of appropriate methodological rigor in their manufacture, they tend to generate either underpre-dictions of future criminality or overpredictions of the same. Moreover, individuals using their own subjective opinions will tend to over- or underpredict criminality. These forms of error are important to understand, and it is in this manner that a false negative or false positive may be generated. The chronic occurrence of these two mistakes in risk prediction can lead to tragic consequences for society or costly expenditures for correctional agencies.

When decision makers are making release decisions for offenders, they will ultimately have to decide if the offender will be allowed out into the community or remain behind bars. There are some implications to these decisions that may not be readily apparent to the casual observer, and these implications, as well as the official professional terms associated with them, should be understood by the student.

First, when correctional decision makers predict that an inmate is not likely to reoffend (and is thus a good risk for community supervision), this is called a prediction in the negative. If in fact the offender is released on community supervision and does not commit any future offense, this is referred to as a true negative. This is because the prediction turns out to be true. Thus, the true negative implies that the offender is predicted to not reoffend and the predic-tion turns out to be true (see Table 3.1 for added clarity). However, if the agency personnel predict that an offender is likewise not likely to offend, but upon community supervision the offender commits some form of crime, this is referred to as a false negative. This is because the agency made the prediction in the negative (meaning it thought the offender would not reof-fend), yet it turned out the prediction was false or incorrect. Thus, the false negative implies that the offender is predicted to not reoffend but the prediction turns out to be false.

On the other hand, if an offender is predicted to reoffend, this is referred to as a prediction in the positive. If the offender is predicted to reoffend but is allowed on community supervi-sion due to the effectiveness of his or her legal representation or some other odd course of events, and he or she then later does in fact commit an offense, this is called a true positive.


Do not

copy

, pos

t, or d

istrib

ute


The true positive implies that the offender is predicted to reoffend, and this prediction later turns out to be true. However, if an offender is likewise predicted to be likely to commit a crime, but later the offender somehow is released into community supervision and is found to never reoffend, this would be a false positive. The prediction was in the positive, indicating risk of reoffending, yet it turned out to be false and thus not accurate.

Naturally, the true positives and true negatives are what agencies hope to obtain as often as possible. These are perfect predictions of offender behavior. However, things not being per-fect, the false positive and false negative predictions are inevitable at some point. Because of this, there are a couple of key points that should be mentioned. First off, false positives are a “safe bet” for correctional prediction boards because if the offender is locked up, the offender simply cannot commit any further crimes in society. From this perspective, it is prudent to simply incarcerate as many offenders as possible (indeed, all of them, optimally speaking) to ensure public safety. However, this would be very costly and can result in excessive expendi-tures on prison systems that have inmates who are neither dangerous nor likely to repeat their criminal behavior. This also can essentially “create” recidivism, as those who would normally refrain from further criminality are placed in an environment where they can learn from and become socialized by other inmates, thereby increasing their incentive to reoffend in the future (students will recall this point made earlier in Chapter 1). Thus, overpredicting offender likeli-hood of recidivating is not a prudent long-term strategy and results in further overcrowding of prisons and ever more costs to taxpayers. By the same token, false negatives result in dangers to public safety and also damage the public perception of community corrections. Ironically, it is often due to the fear of committing a false negative that some correctional agencies may decide to deliberately make more false positives. No criminal justice professional wants to see people in the community get hurt because they commit a false negative, and no agency wants its credibility at risk due to incorrect predictions. Thus, agencies find themselves in a dilemma between the risk of making a false negative and the short-term safety of a false positive.

STATIC AND DYNAMIC RISK FACTORS

Not all risk factors are the same. Indeed, some risk factors are fairly permanent, or at least they occur due to no fault of the offender. Other risk factors, on the other hand, are solely due to the offender. Further, some risk factors are more suited for security, custody, and control of the offender, whereas others are more designed for treatment. Factors such as age at first conviction, gender, sex, and even disabilities or mental impairments are not caused by the offender and are also unlikely to change. These permanent factors are often referred to as static risk factors. Thus, static risk factors are characteristics that are inherent to the offender and are usually perma-nent in nature (Van Voorhis, Braswell, & Lester, 2009). These characteristics are often the best basis for security determinations. Opposite of the static risk factors are the dynamic risk factors. Dynamic risk factors are those characteristics that can change and are more or less influenced or controlled by the offender, such as employment, motivation, drug use, and family relations (Van Voorhis et al., 2009). These characteristics are often most useful to treatment providers

TABLE 3.1 ■ True/False Positives and Negatives in Offender Prediction

Offender Offends Offender Does Not Offend

True Prediction True Positive True Negative

False Prediction False Positive False Negative


Do not

copy

, pos

t, or d

istrib

ute


but are not really a sound basis for security determinations because they have the possibility of changing when certain stimuli are presented in the offender’s life. Figure 3.2 provides an example of a risk-assessment form that incorporates both static and dynamic scales for juvenile sex offenders. The student should note the specific subscales and the separate scoring sections for static and dynamic scores when examining this insert.

FIGURE 3.2 ■ Risk-Assessment Form for Juvenile Sex Offenders

Juvenile Sex Offender Assessment Protocol-II Scoring Form

Scoring Code: 0 = Stable; 1= Moderate; 2 = Severe

Sexual Drive/Preoccupation Scale

1. Prior Legally Charged Sex Offenses 0 1 2

2. Number of Sexual Abuse Victims 0 1 2

3. Male Child Victim 0 1 2

4. Duration of Sex Offense History 0 1 2

5. Degree of Planning in Sexual Offense(s) 0 1 2

6. Sexualized Aggression 0 1 2

7. Sexual Drive and Preoccupation 0 1 2

8. Sexual Victimization History 0 1 2

Sexual Drive/Preoccupation Scale Total: __________

Impulsive/Antisocial Behavior Scale

9. Caregiver Consistency 0 1 2

10. Pervasive Anger 0 1 2

11. School Behavior Problems 0 1 2

12. History of Conduct Disorder 0 1 2

13. Juvenile Antisocial Behavior 0 1 2

14. Ever Charged or Arrested Before Age 16 0 1 2

15. Multiple Types of Offenses 0 1 2

16. History of Physical Assault and/or Exposure to Family Violence 0 1 2

Antisocial Behavior Scale Total: __________

Intervention Scale

17. Accepting Responsibility for Offense(s) 0 1 2

18. Internal Motivation for Change 0 1 2

19. Understands Risk Factors 0 1 2

20. Empathy 0 1 2

21. Remorse and Guilt 0 1 2

22. Cognitive Distortions 0 1 2

23. Quality of Peer Relationships 0 1 2

Intervention Scale Total: __________

(Continued)


Do not

copy

, pos

t, or d

istrib

ute


Community Stability/Adjustment Scale

24. Management of Sexual Urges and Desire 0 1 2

25. Management of Anger 0 1 2

26. Stability of Current Living Situation 0 1 2

27. Stability in School 0 1 2

28. Evidence of Positive Support Systems 0 1 2

Community Stability Scale Total: __________

Juvenile Sex Offender Assessment Protocol-II Summary Form

Static/Historical Scales

1. Sexual Drive/Preoccupation Scale Score:

(Add Items 1–8 [range: 0–16]) __________/16 = __________

2. Impulsive/Antisocial Behavior Scale Score:

(Add Items 9–16 [range: 0–16]) __________/16 = __________

Dynamic Scales

3. Intervention Scale Score:

(Add Items 17–23 [range: 0–14]) __________/14 = __________

4. Community Stability Scale Score:

(Add Items 24–28 [range: 0–10]) __________/10 = __________

Static Score

(Add items 1–16) __________/32 = __________

Dynamic Score

(Add items 17–28) __________/24 = __________

Total J-SOAP Score

(Add items 1–28) __________/56 = __________

Source: This assessment form was adapted from the following U.S. government publication: Prentky, R., & Righthand, S. (2003). Juvenile Sex Offender Assessment Protocol-II (J-SOAP-II) Manual. Washington, DC: Office of Juvenile Delinquency and Prevention. Retrieved from http://nicic.org/Library/019361.


THE APPROPRIATE USE OF SUBJECTIVE AND OBJECTIVE ASSESSMENTS

If the offender is accurately assessed, then only those offenders who are a risk to public safety will be incarcerated. Indeed, those offenders who are a low risk for recidivism, or even those who are a high risk but are most likely to commit nuisance or nonviolent crimes, should not be placed in prison. This is because such forms of custody are very expensive and are designed mainly to prevent the offender from hurting others in society. Despite the common notion that the loss of liberty (the punishment) is the basis for the prison, such a punishment can in fact be fulfilled as well through house arrest or some other form of secure supervision that does not require the expense incurred by prison. Whether the offender is kept within the strict


Do not

copy

, pos

t, or d

istrib

ute


confines of his or her bedroom or the strict confines of a prison cell is not the main concern since both sufficiently restrict liberty. The only difference is that if the offender decides to violate the sentence and leave the confines of his or her bedroom, there is no one immediately there to prevent it, whereas in prison such an option is unlikely due to physical facility con-struction and the immediate reaction of prison staff who are available to respond with haste. Thus, it is the offender’s likelihood to comply with the mandate of the sentence, not the actual restriction of liberty, that should be the criterion for deciding whether to incur the added expense of a prison cell. Otherwise, financial resources are wasted and squandered unnecessar-ily on an offender who could have been just as severely punished for wrongdoings at a fraction of the cost.

It should be noted that assessment is also critical because those offenders who have a high risk of committing violent crimes should not be placed on community supervision. This none-theless occurs due to jail and prison overcrowding, which then places institutions in the pre-carious position of choosing the “least dangerous” of the violent offenders when releasing to the public. This is a very risky position that no administrator relishes, and this is also the most common source of criticism leveled at community release programs. Ideally, these vio-lent offenders should simply not be in the community at all. But given the reality of correc-tions in our society, such offenders should be given supervision that ensures they are under the constant watchful eye of a human supervisory authority so that immediate interventions can be provided just as if they were still within the prison facility itself. Regardless of the cor-rectional system’s situation, it is a strong recommendation of this text that all security-level determinations be based solely on objective assessment instruments and that subjective criteria be avoided when issues of public safety are at stake. This is prudent for agency and personnel liabil-ity purposes, and it provides the most mathematically precise, consistent, and effective means of protecting public safety. On the other hand, when making determinations regarding treat-ment progress, subjective criteria from the specific primary treatment provider should be utilized more than any other form of assessment. These subjective criteria should consist of feedback from both the clinical and the security staff who have had substantial “face time” with the offender. It is important that the student understand the distinctions between the two types of assess-ment and the appropriate times for their use. These two types of assessment will be discussed at length in the following subsections of this chapter. As the student reads through these subsec-tions, it is important to remember their most appropriate use that was just articulated above. This will prove to be a good guiding principle throughout the remainder of the readings for this chapter.

Thus, offender supervision processes revolve around two key forms of response to the offending population. These two responses are simply the goals of (1) incapacitation and (2) treatment. As defined earlier, incapacitation is the process of simply removing the offender from society so that he or she cannot cause further harm to the public. There is no goal beyond this, nor is there any implied retribution or desire to deter other offenders from committing crime. Such goals are too lofty to be pursued with most offenders, and this is especially true with most special needs offenders due to their unique challenges. On the other hand, treatment is the process whereby the offender is provided some form of inter-vention that will help him or her to function within society without resorting to criminal behavior. The student should make one key observation with each of these approaches. Both have the same goal of simply reducing the likelihood of future offending by that specific individual offender. In other words, each of these approaches has the same exact purpose (to prevent the offender from committing future criminal behavior), but each pursues this purpose in a different manner.


Do not

copy

, pos

t, or d

istrib

ute


Subjective Assessment

One of the wisest investments for any correctional system desiring to ensure public safety is in the arena of assessment. More money and resources in assessment mean that the subsequent stages (diagnostic, recidivism prediction, classification) in the equation will also operate bet-ter, resulting in increased public safety as the ultimate answer. One method of assessment, the subjective assessment process of interviewing and observation, is an important yet less struc-tured method of determining the security and treatment needs of the offender. This process entails the use of professionals who utilize their sense of judgment and experience to determine the offender’s possible dangerousness, treatment needs, likelihood of responding to treat-ment, and likelihood of escaping. This is an important process that should not be overlooked. However, it should not be the primary form of assessment but should serve as an integral part of a “two-pronged” assessment process.

There are some drawbacks to this process, just as there are with any form of assessment. For one, the process is subjective, which means that the determination is based upon the impressions of an individual. Thus, these assessments are likely to vary from one professional to another. Second, these assessments can be lengthy in nature and will thus not be useful for large facilities that lack numerous well-qualified staff. Third, these forms of assessment require extensive skill from the staff, and they are therefore only as good as the personnel administering them. Such assessment forms can thus be very costly since only highly educated or well-trained staff will be able to utilize them.

To offset the potential capriciousness of a subjective assessment process, it is suggested that a structured form of interview or observation process be utilized in all cases. A subjective structured interview is simply a process whereby an interviewer will ask a respondent a set of prearranged and open-ended questions so that the interview seems informal in nature (conversa-tional), yet because of the prearranged questions, a structure evolves throughout the conversation that ensures that certain bits of desired data are gathered from the offender. These forms of inter-views are useful since they guide clinicians and other staff who may conduct intakes. Further, the structure of these interviews provides for consistency and uniformity in record keeping, meaning that agencies can ensure similar criteria are considered despite the style and form of the individ-ual interviewer. This provides a certain “baseline” of information against which all offenders are judged. Last, this type of interview ensures that even less qualified staff can conduct an interview and that it will still contain the information considered critical for agency assessment.

Standardized interviews are designed to collect the same type of information from all respondents. If it is a structured interview, all offenders are asked the same questions in the same order, and the answers are recorded in an identical manner (Drummond & Jones, 2010). In unscheduled interviews, the interviewer sometimes varies the sequence of the questions and the order of the topics so that the data can be compared and summarized. The following are some of the advantages and disadvantages of the interviewing process, based on the work of Drummond and Jones (2010).

Advantages:

•• Clients can be guided to answer items completely.

•• Additional information or understanding can be pursued.

•• Nonverbal behavior and cues as well as affective behavior and voice tone and pitch can be observed and noted. Again, this information should be noted only when the nature of the interview allows the clinician to focus on these behaviors as relevant.


Do not

copy

, pos

t, or d

istrib

ute


Disadvantages:

•• Success often depends on the skill of the interviewer to ask the right questions with the right timing and to correctly interpret the observations.

•• The communication of some individuals is inhibited, and clients may simply be unwilling to answer certain questions. They will then either refuse to answer, or just provide a response that is a lie.

•• The personality of the interviewer can greatly influence the outcome.

Drummond and Jones (2010) note that even though the reliability and validity of inter-viewing may be questionable, this is sometimes the only way possible to gather certain types of information. Further, the interviewer can reword questions, add questions, seek clarification of information provided, focus on topics of clinical interest, and so forth. This means that information obtained from such a process will likely be much more relevant to the specific clinical focus, and it also means that helping professionals will have richer information to work with when consulting the client’s case file.

Objective Assessment

As noted earlier in this chapter, when making determinations about security levels (especially when community supervision is involved), it is strongly recommended that determinations be based solely on objective assessment instruments. However, there are a variety of specific assess-ment instruments employed by agencies throughout the United States. These types of objec-tive assessments can range from behavioral checklists that staff complete after a brief period of observation; to paper-and-pencil tests completed by the offender; to assessment formats that characterize the offender’s social, demographic, and criminal history (Van Voorhis et al., 2009). Regardless of the type of objective assessment (and some are better than others), it should always be based upon the response of supervision staff and should never incorporate self-report data from the offender. Further, input from the victim (for example, in programs that utilize a restorative justice element) should not be considered when making decisions to release to the community. Rather, the determination should be clearly restricted to objective mathematical “risk factor” criteria devoid of all other considerations (prison or jail overcrowding, desires of the victim, apparent sincerity of the offender, and so forth). In addition, the type of offender being assessed can be important. For example, Hanson and Morton-Bourgon (2009) compared the accuracy of various approaches to the prediction of recidivism among sex offenders. They conducted a thorough meta-analysis of 536 findings drawn from 118 distinct samples (45,398 sex offend-ers, 16 countries), and empirically derived actuarial measures (objective assessments) were more accurate than unstructured professional judgment (subjective assessments).

Perhaps one of the best-known risk-assessment systems is the Wisconsin Risk Assessment System. During the 1970s, the state of Wisconsin sought to develop a risk-assessment scale that would assess an offender’s likelihood of further unlawful or rule-violating behavior. To achieve this goal, outcome measures were gathered that were based on arrests, misdemeanor convictions, felony convictions, absconding from probation or parole, technical violations, and so forth. After randomly selecting a construction sample, the criminal history and other characteristics of these offenders were entered into a series of statistical analyses to determine the combination of variables that would best predict future behavior of offenders. At the close of this process, 10 specific factors were identified, isolated, and weighted. Many of these same factors appear on most other risk-assessment instruments as well.


Do not

copy

, pos

t, or d

istrib

ute


In addition to these 10 risk factors, this scale includes an administrative policy override. This override consists of an additional factor regarding prior or current assaultive offense his-tory that is weighted in such a manner as to automatically classify an offender as high risk if he or she possesses that factor (Connolly, 2003). These types of overrides are typically policy decisions that are made to classify all offenders with a certain characteristic (such as an assaul-tive offense history) as high risk, regardless of whether the statistical analysis determines that variable to actually be an indicator of likely recidivism. This is designed to address the conse-quences associated with that offender’s potential criminal behavior rather than actual likeli-hood of committing a crime. Thus, if an offender has a stable job with no prior criminal history but he is convicted of a violent sex offense, he may not score high in likelihood of reoffending, but if he were to reoffend, the injury to the community from his crime would be considered too potentially dangerous to place the community at risk (Connolly, 2003).

This form of structured assessment has become the prototype for many probation and parole systems. Staff members use this instrument to score probationers on the predictors contained on the list, and at this point they classify them into high-, medium-, and low-risk categories (Van Voorhis et al., 2009). The items that are included on this list are all statisti-cal predictors of likely failure while on probation. These predictors are all based on previous probation histories among probationers and on the premise that the best predictor of future aggregate probationer behavior is prior aggregate behavior. As a result of tracking probation-ers over time, this instrument has been able to find those factors that are associated with failure and success while on probation. Some of the factors that are examined include the following:

1. Number of address changes in the last 12 months

2. Percentage of time employed in the last 12 months

3. Alcohol consumption problems

4. Other drug consumption problems

5. Offender attitude

6. Age at first conviction

7. Number of prior periods of probation/parole supervision

8. Number of prior probation/parole revocations

9. Number of prior felony convictions

10. Type of convictions or prior adjudications

An illustration of the Wisconsin Risk Assessment scale, as used in the state of California, is provided in Figure 3.3. This example gives the student a clear view of each of the 10 factors used by the model as well as the administrative override variable. Connolly (2003) notes that even though the Wisconsin Risk Assessment is one of the most commonly used instruments among agencies, there is very little comprehensive research that has been conducted on its predictive accuracy. However, based on the research that does exist, there is considerable reason to question the validity of this assessment instrument. The fact that this scale is so widely used but has not been truly validated may be one reason for many of the prediction error rates that have been found to occur throughout the nation. Unfortunately, textbooks on community


Do not

copy

, pos

t, or d

istrib

ute


RISK ASSESSMENT

NAME—LAST, FIRST PROB. NO.X–

DPO CL NO. AO

DATE OF GRANT EXPIRATION DATE OF ASSESSMENT BY

Number of Address Changes in Last 12 Months: _____ 0 None SCORE(Prior to the offense) 2 One 3 Two or more __________

Percentage of Time Employed in Last 12 Months: ____ 0 60% or more(Prior to the offense) 1 40%–59% 2 Under 40% 0 Not applicable __________

Alcohol Usage Problems: _______________________ 0 No interference with functioning(Prior to the offense) 2 Occasional abuse: some disruption

of functioning 4 Frequent abuse: serious disruption:

needs treatment __________

Other Drug Usage Problems: ____________________ 0 No interference with functioning(Prior to the offense) 1 Occasional abuse: some disruption

of functioning 2 Frequent abuse: serious disruption,

needs treatment __________

Attitude: ____________________________________ 0 Motivated to change; receptive to assistance

3 Dependent or unwilling to accept responsibility

5 Rationalizes behavior; negative; not motivated to change __________

Age at First Conviction: ________________________ 0 24 or older(or Juvenile Adjudication) 2 20–23 4 19 or younger __________

Number of Prior Periods ofProbation/Parole Supervision: ___________________ 0 None(Adult or Juvenile) 4 One or more __________

Number of Prior Probation/Parole Revocations: ______ 0 None(Adult or Juvenile) 4 One or more __________

Number of Prior Felony Convictions: ______________ 0 None(or Juvenile Adjudications) 2 One 4 Two or more __________

Convictions or Juvenile Adjudications for: __________ 2 Burglary, theft, auto theft, or (Include current offense.) robbery __________

3 NSF checks or forgery __________ TOTAL __________

Source: Zhan, X., & La Paz, C. (1990). The evaluation of the Wisconsin classification system as it applies to the Los Angeles probation population. Downey, CA: Los Angeles County Probation Department. Retrieved from http://www.nicic.org/pubs/1990/008749.pdf.

FIGURE 3.3 ■ Wisconsin Risk Assessment Scale


Do not

copy

, pos

t, or d

istrib

ute


corrections often present the Wisconsin Risk Assessment scale but do not go into detail as to its validity as an instrument. This is a serious oversight and should be fully addressed within the field of community corrections.

Connolly (2003) conducted one of the most comprehensive examinations of risk-assessment instruments in her dissertation at the University of Texas at Austin. Connolly’s examination was of such caliber as to be funded by the U.S. Department of Justice. In her evaluation of the Wisconsin Risk Assessment scale, she includes research that clearly dem-onstrates the ineffective nature of this instrument. For example, one study by Harris (1994) examined a sample of adult felons on probation. This study sought to compare the predic-tive accuracy of the Wisconsin Risk Assessment with the Client Management Classification (CMC) as well as a combination of the two instruments. Overall, Harris found that the Wisconsin Risk Assessment instrument had high prediction error rates, thus demonstrat-ing this instrument’s lack of validity. In another study, by Yacus (1998), the Wisconsin Risk Assessment scale and the Wisconsin Needs Assessment scale were examined for accuracy in the classification of adult felons in the state of Virginia. This study utilized a sample of 13,011 adult probation and parole offenders who were placed on supervision. Yacus also found high classification error rates for the Wisconsin Risk Assessment. As can be seen in Table 3.2, the results of both studies demonstrate some clear concern for the continued use of this instrument.

When examining Table 3.2, it can be seen that the study by Harris (1994) shows the false positive error rate to be very high (43%), which means the instrument tends to over-predict that offenders will fail on community supervision due to a resulting revocation of their probation or parole status. On the other hand, the Yacus (1998) study examines the likelihood of probation success. In this study, a true positive is generated when an offender is predicted to be successful and turns out to actually be successful in completing his or her probation. Conversely, a true negative for this study occurs when an offender is predicted to fail and does, in fact, fail. From the data in Table 3.2, it can be seen that the Wisconsin Risk Assessment instrument overpredicts that offenders will be successful. Each of these studies examines the instrument from an opposing vantage point. One study examines the instru-ment from the point of predicting that offenders will fail (Harris, 1994), while the other stud-ies the prediction that offenders will succeed (Yacus, 1998). In both cases, the Wisconsin Risk Assessment is found wanting, though of course the results are more significant in the Harris study than in the Yacus study.

Clearly, some serious questions have to be asked in regard to this tool, particularly since its predictive accuracy has not been established. In addition, the overprediction of offending should be considered a serious flaw since this costs needless tax dollars, results in overstuffed prison systems, and is likely to expose offenders to environments that may actually increase the likelihood of reoffending. (As implied in Chapter 1, prison is sometimes equated to an educational camp for career criminals.) One key ingredient to this overprediction

TABLE 3.2 ■ Predictive Accuracy of Wisconsin Risk Assessment (in Percentage)

StudyFalse

PositiveFalse

NegativeTrue

PositiveTrue

Negative Error Rate

Harris (1994) 43 3 13 41 54.5

Yacus (1998) 22 12 57 9 34


Do not

copy

, pos

t, or d

istrib

ute


may be the use of administrative overrides. The use of these mechanisms creates enhanced likelihoods of false positives regardless of what the statistical analyses may otherwise indicate. This may explain much of the inaccuracy involved with the Wisconsin Risk Assessment and indicates that the system might be improved if such mechanisms were more carefully considered.

Another area of concern regarding the Wisconsin Risk Assessment revolves around prob-lems with inter-rater reliability of the instrument. In brief, inter-rater reliability is simply a research term that describes the likelihood that multiple persons rating an instrument will derive similar ratings of that instrument. For a high inter-rater reliability to occur, a larger number of diverse rating experts must have similar judgments in reference to a given variable (such as the likelihood of an offender to continue using drugs). It is clear from past research that the Wisconsin Risk Assessment does not facilitate strong inter-rater reliability and thus indicates that the subjectivity of the rater can negatively impact the outcomes of the instru-ment. Indeed, the question related to the offender’s attitude leaves a great deal to the discretion of the interviewer (ranging from 0 to 5), and attitude itself is a somewhat subjective concept. In addition, questions related to alcohol and drug usage leave room for discretion that can be faulty. Different raters may rate the same offender(s) inconsistently, and this will then weaken the instrument’s predictability (Connolly, 2003).

Since we have determined that reintegrative treatment is a necessary component of any community corrections strategy to reduce recidivism, it stands to reason that variables associ-ated with reintegration must also be considered in the assessment process. In fact, it is often the case that the specific needs of the offender are intertwined with future recidivism. For example, a female offender who is unable to find suitable employment may resort to pros-titution to make ends meet (particularly if she has one or more children to care for). In a similar vein, that same offender may have a drug habit that, when left untreated, can result in further offending to sustain it. Likewise, mentally ill offenders may have medication or treat-ment needs to aid them in maintaining emotional balance. Without such needs being met, the likelihood of future offending is increased. Therefore, when discussing assessment, an additional distinction must be presented. This distinction rests between the understanding of risk-principled assessment and needs-principled assessment of offenders on community supervision.

With risk-principled assessment, the main concern revolves around the protection of society. Within the treatment scheme, the risk-principled assessment system will ensure that hard-core offenders are not in the same treatment regimen as less serious offenders. This may seem to be a topical distinction, but it can be critical to long-term success. For instance, in an anger management program, an offender who has problems with verbal abuse and verbal explosiveness may not benefit from being in an anger management group that has a majority of offenders who are severe domestic batterers. This would be even truer if these assaultive offenders had some form of personality disorder to further interfere with their success. In this case, it is likely that the success of the offender with verbal anger problems might actually be impaired by repeat exposure to these other offenders. These offenders could therefore ruin any positive prognosis for change that the offender might have. In essence, the treatment program could ironically make the offender worse than he was when he started! Though many students may be skeptical, this does in fact frequently happen in many community health service provider agencies. The same could be true in various substance abuse treatment groups and programs and any other treatment regimen that mixes offenders at various levels of therapeutic recovery.


Do not

copy

, pos

t, or d

istrib

ute


Further still, research has shown that intensive correctional treatment programs are more successful with high-risk offenders than with low-risk offenders. Thus, it does not pay to place low-risk offenders in programs designed for high-risk offenders, and this could actu-ally lower their likelihood of success due to negative influence from other high-risk members. Therefore, treatment programs should distinguish between risk principles in their assessment and future placement of offender clients.

Needs-principled assessment, on the other hand, is more concerned with factors spe-cific to the effective treatment and reintegration of the special needs offender. Thus, needs- principled assessment deals with the subjective and objective needs of the offender to maximize his or her potential for social reintegration and to reduce the likelihood of future recidivism. Needs-principled assessment takes into account factors such as substance addic-tion, medical assistance, mental health issues, job development, educational attainment, physical disabilities, and relationships with family or peers. The needs-principled assessment consists mostly of dynamic risk factors, but it will assess offenders based less on whether they possess a certain risk factor and more on the severity of that need or the seriousness of the affliction. For instance, substance abusers whose primary addiction problem revolves around pain relievers may be substantially different from those who abuse to have an elevated mood. Though all are addicts, the type of drug and the type of addiction severity may again warrant differing levels or types of treatment for the offender.

Needs-principled assessment is also multifaceted in nature and goes beyond looking at the main risk characteristic of concern. For instance, a substance-abusing offender may have a wide range of other needs such as the settlement of legal issues, the maintenance of employment, and resolution of family-of-origin conflicts. Each of these issues is not spe-cifically related to the offender’s drug addiction, but each one, if not properly addressed, can impair the offender’s likelihood for further recovery. Rather, the failure to address one of these corollary needs can result in the likelihood of relapse since it is commonly known that drug offenders often resort to drug use during times of stress. Thus, the multifaceted nature of needs-principled assessment involves examining the offender’s global likelihood of reintegration.

Many agencies classify their offenders according to their likely risk of recidivating as well as their identified needs. Aside from the desire to improve the likelihood of reintegra-tion, the use of needs assessments provides a measure to indicate the amount of time and effort that a community supervision officer will spend on a given case in relation to his or her overall caseload. In addition, the use of needs assessments forces qualitative reviews of the offender’s progress that go beyond simply determining if he or she has evaded detec-tion for a technical violation or criminal action. The use of needs-based assessments and resulting classification schemes will be discussed further in Chapter 6, “Needs-Based Case Management and Case Planning,” since needs assessments are directly related to the case-work model of community supervision. As noted in prior chapters, the casework model of supervision is the primary orientation of offender supervision that is supported by this text. Figure 3.4 is provided as an early view of a commonly used needs-assessment scale. This same figure, along with others, will be provided in more detail in Chapter 6. It is hoped that students will keep the current discussion in mind when they begin their reading of Chapter 6, thereby creating a sense of familiarity and continuity in regard to material that is presented throughout the text.


Do not

copy

, pos

t, or d

istrib

ute


FIGURE 3.4 ■ Assessment of Client Needs

(Continued)

Pennsylvania Board of Probationand Parole ASSESSMENT OF CLIENT NEEDS

Select the appropriate answer and enter the associated weight in the score column. Higher numbers indicate more severe problems. Total all scores. If client is to be referred to a community resource or to clinical services, check appropriate referral box.

ACADEMIC/VOCATIONAL SKILLS REFERRAL SCORE

–1 High school or above skill level

–1 Satisfactory employment for one year or longer

–1 Long-standing pattern of self-sufficiency; e.g., good credit rating

–1 Relationships and support exceptionally strong

COMPANIONS

–1 Good support and influence

–2 Exceptionally well adjusted; accepts responsibility for actions

0 Adequate skills; able to handle everyday requirements

0

Secure employment; no difficulties reported; or homemaker, student or retired

0 No current difficulties

0 Relatively stable relationships

0 No adverse relationships

0 No symptoms of emotional instability; appropriate emotional responses

0 No interference with functioning

+2 Low skill level causing minor adjustment problems

+3

Unsatisfactory employment; or unemployed but has adequate job skills

+3 Situational or minor difficulties

+3

Some disorganization or stress but potential for improvement

+2 Associations with occasional negative results

+4 Symptoms limit but do not prohibit adequate functioning; e.g., excessive anxiety

+3 Occasional abuse; some disruption of functioning

+4 Minimal skill level causing serious adjustment problems

+6 Unemployed and virtually unemployable; needs training

+5

Severe difficulties; may include garnishment, bad checks or bankruptcy

+5 Major disorganization or stress

+4 Associations almost completely negative

+7 Symptoms prohibit adequate functioning; e.g., lashes out or retreats into self

+6 Frequent abuse; serious disruption; needs treatment

________ ________

________

________

________

________

________

EMPLOYMENT

FINANCIAL MANAGEMENT

MARITAL/FAMILY RELATIONSHIPS

EMOTIONAL STABILITY

ALCOHOL USAGE

Client Name (Last) (First) (MI) Parole No. SID No.

Release Date (Month, day, year) Agent Name Date


Do not

copy

, pos

t, or d

istrib

ute



–1 Minimum

0 No interference with functioning

0 Able to function independently

0 Sound physical health; seldom ill

0 No apparent dysfunction

0 Constructive activities apparent

0 Low

+3

Occasional substance abuse; some disruption of functioning

+3 Some need for assistance; potential for adequate adjustment; mild retardation

+1 Handicap or illness interferes with functioning on a recurring basis

+3 Real or perceived situational or minor problems

+1 Some constructive activities

+3 Medium

+5

Frequent substance abuse; serious disruption; needs treatment

+6 Deficiencies severely limit independent functioning; moderate retardation

+2 Serious handicap or chronic illness; needs frequent medical care

+5 Real or perceived chronic or severe problems

+2 No constructive leisure-time activities or hobbies

+5 Maximum

MENTAL ABILITY

________

________

________

________

________

________

RECREATION/HOBBY

AGENT’S IMPRESSION OF CLIENTS NEEDS

OTHER DRUG USAGE

HEALTH

SEXUAL BEHAVIOR

Total: _________________

Source: Pennsylvania Board of Probation and Parole website: http://www.pbpp.state.pa.us/portal/server.pt/community/home/5298.

RECIDIVISM PREDICTION

Recidivism prediction is built on the information derived from the presentence interview. It is at this point that correctional personnel attempt to determine the risk involved with allowing the offender to be placed under community supervision. One key quasi-objective clinical inven-tory that is used to determine offender suitability is the Level of Service Inventory–Revised (LSI-R). This inventory was created by Don Andrews and James Bonta, and it has been found to be highly predictive of recidivism among a variety of correctional offender clients (Van Voorhis et al., 2009). The LSI-R is administered by case managers/counselors or mental health pro-fessionals. The assessment process includes a semistructured interview. As with the Wisconsin Risk Assessment System, the LSI-R provides for reassessments of an offender’s risk score. Reassessment of risk scores can be useful when assessing program effectiveness as well as facili-tating program release decisions. In fact, it was found that the LSI-R predecessor, the LSI, was a better predictor of parolee recidivism than the Wisconsin Risk Assessment System.

According to Andrews and Bonta (2003), the LSI-R is a quantitative survey of offender attributes and situations relative to supervision levels and treatment decisions. Designed for offenders age 16 and older, the LSI-R aids in predicting parole outcome, success with offenders


Do not

copy

, pos

t, or d

istrib

ute


in halfway houses and aftercare facilities, and probation recidivism. This inventory consists of 54 items that are based on legal require-ments, and it includes relevant factors needed for making decisions about security risk levels and the likelihood of treatment suc-cess (Andrews & Bonta, 2003). The LSI-R is designed for probation and parole officers, to assist them with decisions about probation and parole placement, security-level classifications, and possible treatment progress.

The LSI-R screening version (LSI-R:SV) consists of eight items selected from the full LSI-R. Like the full version, the LSI-R:SV sam-ples both risk and needs, and the item content reflects four key risk factors: criminal history, criminal attitudes, criminal associates, and antisocial personality pattern (Andrews & Bonta, 2003). In addition, the LSI-R:SV examines other factors such as employment, family, and substance abuse. The items included in the LSI-R:SV not only contribute to its predictive validity, but they also include information that is important to offender treatment planning (Andrews & Bonta, 2003). Each of these items is rated either “yes/no” or “0–3” (0: a very unsatisfactory situation with very clear and strong need for improvement, to 3: a satisfactory situation with little or no need for improvement).

It should be noted that the LSI-R is not intended to replace the professional judgment of the correctional worker. Rather, an objective risk-needs assessment enhances professional judgment, adds to the fairness of offender assessment, and alerts correctional staff to the need for a fuller offender risk-needs assessment (Andrews & Bonta, 2003). Research with the LSI-R shows that scores on the instrument have predicted a variety of outcomes important to offender management. Among probation samples, LSI-R scores have predicted violent recidivism and violations while under community supervision. In assessments of incarcerated offenders, scores have predicted such varied outcomes as success in correctional halfway houses and insti-tutional misconduct (Andrews & Bonta, 2003).

Andrews and Bonta (2003) make it clear that the LSI-R is designed for use as a screen-ing instrument in busy intake settings where, due to time constraints and insufficient staff resources, a complete LSI-R assessment may not be feasible for everyone. The LSI-R also pro-vides a summary of the static and dynamic risk factors that may require further assessment or further intervention from agency personnel.

In addition, the LSI-R fits a rather specific type of treatment model. As a risk-prediction inventory, the LSI-R fits best with programs that are based on clear cognitive-behavioral and social learning treatment modalities. This is not a problem, however, because most treatment programs in the criminal justice system are based on such orientations. But if clinicians desire an accurate assessment of the offender’s likelihood of reforming, they must keep in mind that the LSI-R has limits that are grounded in cognitive-behavioral approaches. Van Voorhis et al. (2009) point out that most research on cognitive-behavioral and social learning approaches shows that this modality is the most effective, overall, when dealing with the offender popula-tion. Thus, the LSI-R is ideally suited and designed for those programs that utilize the most effective modalities. The LSI-R therefore dovetails nicely with these programs and lends further validity to the nature and intent of therapeutic treatment programs in the criminal justice field.

PHOTO 3.2 This laptop has GPS tracking software loaded and is used to keep tabs on offenders of different classification levels throughout the local community. In addition to classification information, this software can determine the area of the city in which the offender is located (the zone), whether the offender has a curfew requirement, and whether he or she is required to wear a bracelet strap. All of these aspects (as well as other custody and security features) are noted at the bottom of the screen.


Do not

copy

, pos

t, or d

istrib

ute


THE LINK BETWEEN THEORY AND RISK PREDICTION

Connolly (2003) notes that in order for significant improvements in risk prediction to occur, instruments will need to be better grounded in theory. Specifically, actuarial prediction models to be used with serious offenders should be directly based on the discoveries made by research that empirically tests the ability of theories to explain and predict crime. Many students fail to appreciate the important connection between theoretical explanation of crime and our ability to predict crime. In essence, theory is nothing less than a macro-level form of assessment. What many laypersons and even experts in the field may not realize is that many variables used to predict criminal behavior in theoretical tests are identical to those used in assessment instruments. Thus, there is a direct relationship between the research of theo-rists and the potential for improvement of risk-assessment instruments. Indeed, as theoretical research continues, fertile ground is created for further refinement of instruments that use the same or similar variables. However, most current models are based on their simple ability to predict the likely outcome, rather than their ability to explain the reason for criminal or non-criminal behavior (Connolly, 2003). This is problematic, particularly when one considers that needs-assessment variables are more related to explaining why recidivism might occur, and these variables are critical to the ultimate success of offender reintegration. Consider the point made by Krauss, Sales, Becker, and Figueredo (2000), who state,

Probation risk assessment and other forms of risk assessment have become exclusively based on prediction rather than explanation of behavior. Actuarial assessment instruments are, by and large, atheoretical, and consequently, do not effectively examine the causes of the behavior that the instruments are designed to predict. Present methods of probation risk assessment simply highlight individuals who are high risks for recidivism, without explaining why these individuals are more likely to recidivate. (p. 92)

Explaining the “why” behind criminal behavior can improve the treatment prognosis for most offenders (incorporating a needs-based approach that is best addressed through a case-work model), which in turn improves their likely reintegration. While there is no single best theory to explain all crime, there are some common starting points that lend themselves well to risk prediction. When considering these starting points, Connolly (2003) notes that the most promising research with direct application to the construction of adult offender risk prediction is based on one of the following three criminological theories: the general theory of crime, or self-control theory, by Gottfredson and Hirschi (1990); the age-graded theory of informal social control by Sampson and Laub (1993); and social learning theory by Ronald Akers (1985).

Gottfredson and Hirschi’s (1990) self-control model suggests that people differ in their simple will to refrain from criminal behavior. Simply put, a more disciplined person will be less likely to commit crime. Correspondingly, individuals with low self-control are more likely to exhibit criminal behaviors. One application of Gottfredson and Hirschi’s self-control the-ory to predicting adult criminal offending used a six-factor, 24-item scale that had been used in various studies with a variety of subpopulations (Connolly, 2003). A comprehensive test of this scale was conducted on a cross-cultural sample of 8,417 adolescents from four different countries (Vazsonyi, Pickering, Junger, & Hessing, 2001), who were administered the 24-item scale. One of the significant findings of the study is that the scale is predictive of deviance, with a risk-seeking or high sensation–seeking life view explaining up to 25 percent of the potential causes of continued criminal behavior. Connolly notes that the scale has performed well in


Do not

copy

, pos

t, or d

istrib

ute


numerous other studies as well. For instance, Alarid, Burton, and Cullen (2000) tested five measures of social control: marital attachment, attachment to parents, attachment to friends, involvement, and belief. The social control variables of attachment to parents, involvement in conventional activities, and belief in the law were all significantly correlated with future criminal behavior. Further, these predictors could be pinpointed to the specific type of crime that is likely for a given offender. For example, marital attachment only impacted and reduced the likely involvement in property crimes, while attachment to peers was positively correlated with criminal activity.

Sampson and Laub’s (1993) theory extends the concepts of social control into adulthood. The authors emphasize the quality and nature of informal social ties rather than simply noting that such exist. For example, merely being employed is not necessarily as effective an indicator as knowledge of the offender’s desire to keep that job. Connolly (2003) points out that “adult risk prediction instruments frequently include the basic construct of social control variables, however, they fail to go to the next step in measuring the quality and strength of the social ties” (p. 156). Rather than simply noting if an offender has a job, it may be equally important to know about the stability of the job (seasonal, temp, and so forth), the type of work, and the work habits of the offender. Such factors have been found to be important in predicting future recidivism, with low job stability making an offender 5 times more likely to carry out a future criminal behavior (Connolly, 2003).

Finally, social learning theory has not been as widely tested in recent years as self-control theory. However, Alarid et al. (2000) tested the correlation among individual definitions of crime, others’ definition of crime, and criminal friends with the likelihood of criminal behav-ior. All three of these associations were found to be statistically significant, directly correlating with drug and property offenses as well as violent offenses. As has been noted in earlier chap-ters of this text, social learning theory is the primary theoretical orientation that is thought to extend the concepts of the classical school of criminology. In addition, this theoretical per-spective lends itself well to the social casework model that is a primary theme of Chapter 6, later in this book.

The previous research demonstrates the usefulness of theory and its connection to risk prediction. “Constructing adult risk prediction models that are grounded in theory is a direc-tion the field of criminology needs to take if any substantive gains in predictive power of these instruments are likely” (Connolly, 2003, p. 156). Researchers and agencies wishing to improve adult risk-prediction scales would thus be wise to incorporate the quantitative research that has proliferated within the field of criminological theory.

APPLIED THEORY

Criminological Theory and Risk Prediction

Though this chapter makes it clear that theory aids in explaining the “why” behind criminal offending, students should understand that the “why” also aids in predict-ing future criminal behavior. This is an important point because many people may not be aware that crimino-logical research examines variables that are quantified,

measured, and examined for significance in likely criminal behavior. This is nothing less than a form of prediction.

The primary difference between purely theoreti-cal research and pure risk prediction is that theoretical research starts with a hypothesis that is designed to test the efficacy of a theory in explaining criminal behavior while, on the other hand, risk prediction presumes these variable-based hypotheses are correct and uses them to

(Continued)


Do not

copy

, pos

t, or d

istrib

ute


BETTER DIAGNOSIS: THE NEED FOR IMPROVED ASSESSMENT

The information for this section is drawn largely from the work of Tony Fabelo and Geraldine Nagy (2006), two widely respected authorities on correctional research in the state of Texas. Much of the material supporting this section was produced by the JFA Institute in Washington, D.C., and is derived from the federal publication titled Better Diagnosis: The First Step to Improve Probation Supervision Strategies. As the authors of that publication, Fabelo and Nagy note that without a diagnosis of offenders along risk and criminogenic factors,

it is very difficult to: (a) distinguish offenders along characteristics that identify their supervision needs; (b) guide judges in setting appropriate conditions of supervision; (c) guide probation administrators in designing differentiated supervision strategies; (d) provide probation officers with reliable information to formulate and implement effective supervision plans; and (e) devise clear outcome expectations for different populations. (p. 4)

Though this statement may seem fairly straightforward and deceptively simple, it fully captures the essence and importance of assessment.

Fabelo and Nagy (2006) conducted a comprehensive examination of assessment processes in Travis County, Texas (the county that includes the state’s capital, Austin). They found that the assessment processes used at that time faced the following issues:

1. PSI officers did not develop PSIs that followed a comprehensive, standardized interview protocol, nor did they integrate the results from a risk assessment and case classification into their final diagnosis. The PSIs were essentially “biographies” characterized by long narratives developed by the PSI officers following a general interview guideline. It was found that these narratives were potentially affected by the different writing styles and predispositions of the officers when interpreting the offenders’ responses to open-ended interviews.

2. Assessment tools were in place but were not properly used. For example, the risk assessment in that county was typically done after the PSI was completed by the

(Continued)

predict the likelihood of criminal behavior in the future. It is in this manner that the two are interconnected, one with the other, in a circular relationship. Indeed, as practitio-ners use risk-prediction devices, the effectiveness of these instruments will be established and, over time, will validate (or disprove) the theoretical bases that were initially used during their construction.

Risk-prediction instruments are the practical outcome of widespread criminological and psychometric research.

In other words, these instruments tend to combine find-ings from a wide range of theoretical approaches, with the single-minded purpose of predicting recidivism. Thus, these instruments are a composite of a variety of theoreti-cal components, ideally mixing and matching various con-structs from multiple theoretical perspectives in a manner that optimizes prediction (leading to true positives and true negatives) and minimizes the likelihood of error (false neg-atives and false positives) in the risk-assessment process.

Source: Lilly, J. R., Cullen, F. T., & Ball, R. A. (2007). Criminological theory: Context and consequences (4th ed.). Thousand Oaks, CA: Sage.


Do not

copy

, pos

t, or d

istrib

ute


assigned field officer during an “initial interview” with the offender. Consequently, risk levels were not considered during the setting of conditions or the assignment of cases.

3. The PSI officer would make recommendations to the judge, but in many cases the conditions of probation were not well coupled with the supervision requirements.

4. A large number of requests were observed to modify the conditions of supervision after an offender was placed on probation. This was interpreted as being the result of not having a clear diagnosis of the offender at the point of adjudication, thereby undermining the match of the conditions of supervision to the appropriate supervision strategies.

5. It was found that the paperwork to request supervision modifications required a substantial amount of time and effort. This was in addition to the roughly 8 to 10 hours to complete the general PSI. Further, other assessments such as those associated with substance-abusing offenders took another 2 to 3 hours. (As has already been noted in Chapter 2 and as will be discussed in Chapter 4, the excessive paperwork associated with the community supervision officer’s job causes a great deal of stress and generates inefficiency in agencies.) Moreover, these assessments often required multiple visits by the offender to the agency before they were completed. This tended to create additional burdens on community supervision staff as well as on other administrative personnel.

Figure 3.5 summarizes the main issues that were addressed during Travis County’s reform process to rectify problems with its assessment process. The five previous observations from Fabelo and Nagy (2006) tie in with this illustration. The assessment process and challenges with the integration of the PSI process are illustrated in Figure 3.5.

From the noted deficiencies in the assessment process used by Travis County, it became clear that the assessment, diagnosis, and classification process would require substantial improvements, particularly in regard to the use of the PSI. Through the use of a commit-tee of probation experts, officials, and seasoned practitioners, a number of improvements were generated to the process. It was found that a wide range of forms were used as part of the assessment process. In fact, nearly 40 forms were used, and this entailed nearly 100 pages of content when complet-ing the process. Figure 3.6 illustrates the number and type of forms that were used by the Travis County system, and this was also reflective of what many other counties in Texas used. These forms were categorized as (a) interviews, question-naires, and data collection; (b) assessments and screenings; (c) movements and referrals; (d) consent forms; and (e) ending evaluations and exit reports.

When examining this process, experts in Travis County observed that much of this paperwork either was not related to the overall cohesive program of community supervision or duplicated the information included on other forms. In essence, much of the paperwork was either not necessary at all or redundant. Further, it was found that many of the forms could easily be collapsed down into single forms of combined administra-tion. This is actually a very important point from both an assessment and an individual practi-tioner perspective. Further elaboration is necessary to clarify both of these points.

PHOTO 3.3 An intake and classification worker enters information that will aid in determining an offender’s eventual classification level, both in the jail and later on community supervision.


Do not

copy

, pos

t, or d

istrib

ute


FIGURE 3.5 ■ Assessment Process and Key Issues Addressed in Travis County, Texas

Source: Adapted from Fabelo, T., & Nagy, G. (2006). Better diagnosis: The first step to improve probation supervision strategies. Washington, DC: JFA Institute.

1. PSI narrative format provides for inconsistent assessmentsand reporting; does not rely on well-structured interviewprotocol; takes from 8 to 10 hours per offender to complete

2. Court officialsreceive a PSI

“biography” of theoffender and not adiagnosis driven

by evidence-basedtools

3. PSI does not relateto the setting of

supervision strategies;by policy, the

department overridesrisk-assessment

results; officers do notuse SCS tool for low

and medium cases andfor few maximum cases

as requiredSupervision Plan

Developed

First ProbationOffice Visit

Court DispositionOffender granted

probation

Intake Section“Admits” offenderinto department

4. Substance abuseassessment is a

separate process;process takes another

2 to 3 hours; datacollection is duplicated;

not all offenders areassessed during PSI;

not all felony offendersare routinely assessedfor substance abuse

TAIP substanceabuse assessment

PresentenceInvestigation

Offender in jail orout on bond awaiting

disposition

First, the use of excessive paperwork creates a disorganized form of offender assessment. What may seem to be comprehensive is, in actuality, overstuffed with information that may not even be closely examined when it is understood by agency personnel that forms are redun-dant or unrelated to the true process of setting a supervision strategy. Such a system detracts from the information that is included, since this information gets buried in the redundancy or lack of pertinence. If this occurs, it can easily lead to human error when already overworked community supervision staff are expected (unrealistically) to incorporate all of these sundry details into a genuine assessment. While a comprehensive report is indeed desired, it is impor-tant that agencies not increase the strain on already tired and sore eyes that must pore through volumes of information that may, in the real world of supervision, simply get lost in the shuffle. Thus, all paperwork should be relevant to primary concerns with supervision and treatment strategies and should be centralized to key forms as much as possible. This will improve the use


Do not

copy

, pos

t, or d

istrib

ute


of the information that is collected, thereby contributing to an enhanced sense of efficiency and improved public safety.

Second, it will be seen in Chapter 4 that excessive paperwork happens to be one of the most common sources of stress and burnout among community supervision officers. This problem should not be taken lightly. Keeping knowledgeable and committed staff within an agency is critical to that agency’s performance and is therefore critical to the community’s safety. If com-munity supervision officers are to be expected to maintain human face-to-face contact with their offender caseload, they must not be burdened with paperwork that is irrelevant or redun-dant. Otherwise, agencies will continue to grapple with turnover due to job stress (often cited as being caused by paperwork expectations that compete with time demands from other aspects of the job). As will be seen in Chapter 4, this is an often-cited issue of frustration for community supervision officers and is therefore worthy of serious consideration.

FIGURE 3.6 ■ Forms Used in Travis County PSI and Intake Process

Source: Adapted from Fabelo, T., & Nagy, G. (2006). Better diagnosis: The first step to improve probation supervision strategies. Washington, DC: JFA Institute.

Interviews,Questionnaires,Data Collection

Assessments andScreenings

Movements andReferrals

Diagnostic Presentence Information

Intake Setup Form

Victim Interview Guide

Caseload Assessment Form

Vendor Financial Study

Staffing Sheet

Initial Interview Questionnaire

SMART Psycho-social Interview

Substance Abuse Evaluation (SAE)

SASSI Screening

Adult Placement Education Tool

Strategies for CommunitySupervision (SCS)

Wisconsin Risk Assessment

Static 99 Assessment for SexOffenders

Internal

Felony Court

Referral

Diagnostic Assessment

Checklist

Intake Assignment Form

Missed Appointment Notice

General Orientation to Comm Supv

Pre-employment Program Form

Things to Bring to an Initial CSCD General Referral Form

External

Request to Reset Sentencing Date

Wishing to Proceed Out of State

Ignition Interlock

Consent Forms

Release of Confidential Info

TAIP Consent Forms

TAIP “No Harm” Client Contract

SMART Consent Forms

Travis County Consent Forms

End ProductReports

Presentence Investigation Report

Risk Assessment (Perkins only)

Supplemental Information

TAIP Admission Approval Form

Recommended Conditions of Probation

Offender Supervision Plan


Do not

copy

, pos

t, or d

istrib

ute


CROSS-NATIONAL PERSPECTIVE

The Adult Actuarial Risk Instrument (AARI)—A Model Risk-Assessment Instrument From Australia

Max Maller of the Crime Research Centre of the University of Western Australia and Richard Lane of the Department of Justice of Western Australia presented a paper high-lighting a newly developed risk-assessment system. This system was named the Adult Actuarial Risk Instrument (AARI), and this computer-aided program was developed over five years by the Crime Research Centre. The AARI was designed using data that spanned 16 years, from 1984 to 2000. The data used when norming and design-ing this instrument consisted of over 620,000 arrests that included dynamic and static risk factor information. Information such as sex, arrest cardinality (position in the offender’s arrest career), age at arrest, race, and most serious offense are some examples of the com-mon data utilized. The AARI system continually iterated over time through a process that includes new data and archiving of outdated data. In addition, the AARI included a “Case Needs” assessment that contained ratings for eight domains: (1) occupation, (2) marital/family, (3) asso-ciates/social interaction, (4) alcohol use, (5) substance use, (6) community functioning, (7) personal/emotional orientation, and (8) attitude.

The AARI was created to better identify high-risk offenders. Previous risk-assessment tools in Australia had problems because the data chosen had only dealt with a small group of offenders during a particular time, thereby lacking external validity. Further, many of these instruments were not always reliable over a longer period of time. The AARI is an example of how actuarial models are used throughout the world to predict offender recidi-vism. Specifically, the student should note the following points regarding this assessment instrument developed in Australia:

1. Maller and Lane (2002) refer to the Wisconsin Risk Assessment System and the Canadian Level of Service Inventory when discussing the creation of the AARI. This demonstrates that researchers and practitioners around the world face similar problems and that, as a whole, they tend to learn from one another even if oceans stand between them.

2. The AARI is developed from the integration of police data, prison data, and community supervision data or information. This means that there are multiple data set observations of the offender and, thus, the data are likely to be more valid in determining risk of offenders since there are multiple points of observation within the justice system.

3. Maller and Lane (2002) note that the AARI accounts for the fact that substantial differences in rate and type of offending occur between male and female offenders. This naturally reiterates the points made earlier in this chapter by Austin (2006), who noted that these two populations should not be assessed similarly.

4. The AARI uses a needs-based assessment as well as the typical variables used to predict recidivism. As Maller and Lane (2002) point out, this distinguishes “between risk control—those constraints and sanctions placed on an offender to ensure that he/she remains less likely to reoffend while under supervision; and risk reduction—the effort to achieve a permanent reduction in the offender’s likelihood of reoffending” (p. 2). Thus, needs-based variables aid in allocating supervision resources so as to reduce likely future offending, rather than simply predicting it. It should be noted that though the correlation between actuarial risk and “Case Needs” scores was not strong, a positive correlation between the two was nonetheless found to exist.

The AARI has proven to be highly accurate, and the “Case Needs” inclusion has helped prioritize those situations where changes are needed, even with individual offender cases. This instrument is also useful as a case manage-ment planning tool, and is thus able to enhance reintegra-tive efforts among those individuals who are most likely to reoffend. Students are encouraged to read the entire arti-cle by Maller and Lane (2002) found at http://www.aic.gov .au/media_library/conferences/probation/maller.pdf, since this instrument is an excellent cross-national example of both a recidivism prediction instrument and a “Case Needs” assessment instrument.

Source: Maller, M., & Lane, R. (2002). A risk assessment model for offender management. Paper presented at the Probation and Commu-nity Corrections: Making the Community Safer Conference convened by the Australian Institute of Criminology and the Probation and Community Corrections Officers’ Association Inc., Perth, Australia. Retrieved from http://www.aic.gov.au/media_library/conferences/probation/maller.pdf.


Do not

copy

, pos

t, or d

istrib

ute


CLASSIFICATION

As has been addressed in earlier chapters, the presumption of this text is that recidivism can be more effectively reduced by ensuring that reintegrative efforts are maximized for offenders in the community. As was just pointed out, risk assessment of offenders is often simultaneously tied to their specific treatment needs. This means that security and treatment needs are not always easy to untangle from one another. Because of this, agencies must not only determine the general risk level of the inmate or offender on community supervision, but must also correctly “match up” the offender’s treatment plan with the level of supervision determined by the LSI-R and other risk-prediction tests and procedures. One primary tool used in corrections is an instrument known as the MMPI-2, which is the most widely used objective test instrument in corrections. The MMPI-2 is an objective personality adjustment inventory test that can be given to large numbers of offenders at the same time or individu-ally as desired. The MMPI-2 has 567 true/false questions that require the offender to be able to read at the sixth-grade level. Further, the MMPI-2 has been restandardized and is on tape for individuals who are blind, illiterate, semiliterate, or disabled. It is important to stress that the MMPI-2 is primarily a clinical tool used for detecting mental health disorders among abnormal populations.

This test has a number of “subscales” within it. These subscales are a series of questions that are embedded and camouflaged within the 567 total questions and are dispersed at random points within the test. The questions are all designed to measure specific points of interest, to provide a multiprofile view of the offender’s personality. However, the MMPI-2 is very effective with the manipulative offender population because of three specific subscales that are included. These subscales of interest are the Lie (L), Infrequency (I), and Correction (K) scales.

The Lie or “L” scale consists of 15 items (out of 567 total) such as “I never get angry.” The scale indicates whether the client is consciously or unconsciously presenting him- or herself as a perfectionist. The “I” scale consists of 64 questions (again, out of 567) but does not measure a trait. Because the items are answered in a deviant direction by less than 10 percent of those who take the test, a high score indicates that the offender has endorsed a large number of serious psy-chological items. For offenders, this may be an attempt to look bad on the test, or they may be confused or even having delusions (this is especially likely with offenders who are mentally ill).

Compare and Contrast Exercise

1. As can be seen from reading this Cross-National Perspective, there are very many similarities in how risk-assessment is performed in the United States and Australia. Identify and discuss three of these similarities that were identified in this reading.

2. Discuss some of the other uses of the AARI beyond simple risk prediction. Are these additional uses

commonly also integrated with models used in the United States?

3. What are some of the perceived benefits to having a more accurate risk-assessment instrument? What are some of the problems?

4. Does the “Case Needs” supplement sufficiently address external factors in risk assessment?


Do not

copy

, pos

t, or d

istrib

ute


Last, the “K” or Correction scale measures defensiveness as a test-taking attitude. The scale has 30 items that cover a wide range of content areas. Low scores usually indicate a deliberate attempt to appear bad, but sometimes a self-critical offender (an addict, some pedophiles, or an offender who feels remorse) may endorse responses that indicate pathological tendencies.

In addition to the MMPI-2, the MMPI-2 Criminal Justice and Correctional Report is based on decades of research and is designed to more closely fit the outcome data from the MMPI-2 to a classification scheme (Megargee, 2005). The MMPI-2 Criminal Justice and Correctional Report (MMPI-2 CJCR) is perfectly suited to “match up” the offender’s treat-ment plan with the level of security and serves as an additional double check when making security decisions from the LSI-R. This report is used in conjunction with the MMPI-2 to provide information pertaining to the offender’s needs assessment, risk assessment, and pro-gram planning within a correctional agency. The report is designed to identify those offenders who may suffer from thought disorders, serious depression, and substance abuse problems. It identifies those who may need mental health treatment as well as those who are most likely to be hostile, predatory, bullied, or victimized while incarcerated. This report also includes predictor items related to self-injury and suicide.

The MMPI-2 CJCR system was developed by Edwin Megargee and is fashioned around the well-regarded Megargee Offender Classification System. This system of classification is known to provide solid empirical support for classification and placement decisions. The Megargee system is especially effective in assisting criminal justice practitioners to deal with an offender population that is increasingly including the mentally ill or disordered offenders within its ranks. Further, the Megargee system has been reported to effectively classify 90 to 95 percent of the MMPI-2 profiles encountered in most probation, parole, and institu-tional correctional settings. The Megargee Offender Classification System is the basis for the MMPI-2 CJCR. The results from this report can be used to support important management, treatment, and programming decisions. This includes the ability to do the following:

•• Reliably classify offenders at initial intake of incarceration to support important supervision and treatment issues

•• Identify offenders who may present less risk to the system, possibly allowing a downgrade in security level and improvement in placement on community supervision

•• Better understand an offender’s background, attitudes, and abilities to determine if the individual will benefit from substance abuse treatment, mental health programming, and other services

•• Identify offenders who may do well in prison work programs based on their educational and vocational abilities

•• Address readiness of offenders to leave the institution and to assist in developing effective aftercare programs

•• Accurately evaluate and reclassify offenders over the course of their supervision to support programming or treatment decisions

The MMPI-2 CJCR also consists of nine behavioral dimensions that compare offenders to other offenders rather than to the general outside population to ensure that results are correctly normed. The following is a list of these behavioral dimensions:


Do not

copy

, pos

t, or d

istrib

ute


1. Apparent need for further mental health assessment or programming

2. Apparent leadership ability, dominance

3. Indications of conflicts with or resentment of authorities

4. Likelihood of positive or favorable response to academic programming

5. Indications of socially deviant behavior or attitudes

6. Apparent need for social participation, extroversion

7. Likelihood of mature, responsible behavior; positive response to supervision

8. Likelihood of positive favorable response to vocational programming

9. Likelihood of hostile or antagonistic peer relations

Further, the MMPI-2 CJCR identifies nine possible areas relevant to the offender. This provides treatment staff with indicators of difficulties that the offender may face. As with the other report components, the offender’s problems are normed against a population of other offenders to ensure that comparisons are similar and that the test is valid for the offender population. The nine problem areas identified are as follows:

1. Difficulties with alcohol or other substance abuse

2. Manipulation or exploitation of others

3. Thought disorders

4. Overcontrolled hostility

5. Family conflict or alienation from family

6. Depressive affect or mood disorder

7. Awkward or difficult interpersonal relationships

8. Anger control problems

9. Tendency to get sick/ill frequently

When taken together, the MMPI-2 and the MMPI-2 CJCR provide a comprehensive means of classifying offenders based on both mental health and criminal justice categories of concern. The use of the LSI-R serves as an initial predictor of recidivism, and it is desir-able to have one single instrument strictly for this purpose. The MMPI-2 and the MMPI-2 CJCR both go beyond mere recidivism prediction by including mental health and secu-rity classification determinations. The process described in this chapter would provide for optimal assessment, security, and treatment of offenders placed on community supervision. Simultaneously, the offenders would gain excellent treatment programming from a system that would address specific areas identified as possible “problem areas” in their effective reintegration.


Do not

copy

, pos

t, or d

istrib

ute


CONCLUSION

The collection of information is an important first step in the intake and assessment process. It is critical that community supervision officers ensure that informa-tion included in the PSI is accurate, since this document affects sentencing, later supervision levels, later treatment program participation, and release decisions that fol-low. In fact, it could be said that the PSI is a combined subjective and objective assessment document, and it is therefore fitting that it is discussed from an assessment or classification perspective.

Both subjective and objective assessments should be used when making security and treatment decisions for special needs offenders. However, security decisions should be based solely on objective risk-prediction cri-teria, while treatment decisions should be based more upon subjective structured interviews and clinical diagnoses so as to capitalize upon the expertise of the clinician. The Wisconsin Risk Assessment System was presented as the recommended objective assessment scale for security-based decisions, while the Level of Service Inventory–Revised (LSI-R) was presented as

the preferred instrument at the recidivism-prediction stage due to its utility with large populations, its quan-titative aspect, the ease with which practitioners may use the instrument, and its ability to directly link with programs based on cognitive-behavioral treatment. The problems and pitfalls associated with false positives and false negatives were noted to provide a general overview of the problems associated with inaccurate assessment in correctional systems. This demonstrates that it is assess-ment that provides the basic building blocks to the suc-cess of community corrections as a whole. The MMPI-2 and the MMPI-2 Criminal Justice and Correctional Report were presented as the premier tools for effective classification of offenders based on their self-reported pathology and a variety of behavioral dimensions and problem areas that were identified. These tools are spe-cifically designed to be consistent with the Megargee Offender Classification System, which was presented as a top-quality classification system for utilization by any agency that has a substantial population of special needs offenders.

KEY TERMS

Dynamic risk factors 64False negative 63False positive 64Level of Service Inventory–

Revised (LSI-R) 76Megargee Offender Classification

System 86

MMPI-2 85MMPI-2 Criminal Justice and

Correctional Report 86Needs-principled assessment 74Presentence investigation report 52Prognosis 52Risk-principled assessment 73

Static risk factors 64Subjective assessment process 68Subjective structured interview 68True negative 63True positive 64Wisconsin Risk Assessment

System 69

END-OF-CHAPTER REVIEW: SHARING YOUR OPINION

1 Understand the importance of the PSI and why accurate information is critical to later supervision and treatment needs of the offender.

2. In your opinion, do most practitioners actually fully utilize assessment processes in their day-to-day operations?

3. In your opinion, would most practitioners really be willing to avoid false positives when blame can be

pointed at them should an offender actually turn out to be a false negative?

4. Identify the flaws with the Wisconsin Risk Assessment and explain whether these flaws seem serious or more of a nuisance.

5. In your opinion, what are some of the key strengths in using the MMPI-2 Criminal Justice and Correctional Report?


Do not

copy

, pos

t, or d

istrib

ute


“WHAT WOULD YOU DO?”

You are the chief probation officer in a small community supervision agency in the rural part of your state. Your own particular state does not centralize probation ser-vices. Instead, each county is responsible for funding and implementing its own community supervision programs. You have talked with the chief judge of the county, and you both have decided to implement a more standardized form of offender classification within the agency. The chief judge explains that she wants you to spearhead the design of such a classification system.

The chief judge explains that money is limited but some is available. Your task is to create an assessment and classi-fication system for your agency that is based on elements of other standardized assessment or classification tools that exist. Further, the chief judge says that she can probably get funding for at least one scale of your choosing, but you will need to determine the price for that scale. Thus, your task is to develop an assessment and classification system for your agency and identify one scale that you would purchase to augment your own agency-made system of assessment.

For this project, you are to conduct online searches and find one of the standardized scales mentioned in this chapter. You must find the company that produces that scale and also find the cost of that scale on the open market. Be sure to cite and reference your source in APA format. Next, you are to outline the steps that you would take to produce your own agency-created assessment instruments. Since you are a chief proba-tion officer, it is a good thing that you read current lit-erature on community supervision issues. As such, you just completed reading this chapter of your textbook, and you may use the information from this chapter to guide you in completing your task, but you should be specific when outlining your plan. Further, the chief judge wants you to explain why you would choose those particular processes. With that in mind, you must clearly detail the process and explain the actions that you decide to take.

What would you do?

APPLIED EXERCISE

Complete the following questions as if you were con-structing a PSI for Juan:

Juan is a Mexican American youth who has just turned 17 years of age. He lives in Los Angeles with his mother and two brothers. Juan’s father is currently in prison for trafficking drugs and is not expected to be released for several more years. Juan has recently been convicted of committing simple battery of another youth. Police information indicates that Juan is suspected of being a member of a local street gang, and police believe that Juan’s offense was gang related.

This is Juan’s first offense, and Juan does not have the appearance of being a hardened youthful criminal. Further, Juan is known to occasionally talk with Father Miguel, a local priest of the Catholic Church in the area. Juan has been noted to respect Father Miguel’s guidance and advice; however, Juan’s brothers do not hold similar respect for the Church. In fact, one of Juan’s brothers is a known gang member.

At school, Juan is basically a quiet teen; he has a girl-friend, Miranda, who was expelled from school at one point due to drug use on school grounds. Juan has been

labeled as having a math and reading learning disability by the school system. He is fully bilingual but does not excel at school regardless of whether the curriculum is in Spanish or English.

Juan’s mother works hard at a local grocery store. She comes from the area, and, upon asking, you learn that most of the men in her family of origin were or currently are affiliated with the Mexican Mafia. Juan’s mother, however, has never had any official contact with the criminal justice system and has a completely clean record. She attends the same church that Father Miguel presides over, and she is also active with a variety of church groups.

Students should complete this application exercise as a mini paper that explains the scenario and then addresses each question throughout. Total word count: 1,200 to 2,000 words.

Students: You know that there are many other things that you need to include in Juan’s PSI, and you know that the information above is important. With this in mind, answer the following questions:


Do not

copy

, pos

t, or d

istrib

ute


1. What points of information do you think are most critical from the information noted above?

2. Who would you talk with first when conducting the PSI? Who would be second and third on your list? Explain why.

3. What other issues would you seek to determine when constructing the PSI? How would you go about determining them?

4. How likely do you believe it is that Juan will recidivate?

5. Would you recommend Juan for probation or some sort of institutionalization? Explain your answer.

FOOD FOR THOUGHT

As previously noted, the LSI-R has been validated for a number of populations including men, women, juveniles, and ethnic minorities. One area that has not been specifi-cally examined is the appropriate use of this instrument for drug-involved offenders. For this article, Kelly and Welsh (2008) expanded efforts to validate the use of the LSI-R for 276 drug-involved offenders using a bivariate correlation and logistic regression analysis to examine the

appropriateness of this instrument for this population of offenders. Results of this study suggest that the LSI-R is a “stable, significant predictor of reincarceration” (p. 828). The authors further explored the relative impact of the drug and alcohol problem subscale use of this instru-ment. This exploratory study sets the foundation for further examination and validation for the use of the LSI-R on specific subgroups.

The Predictive Validity of the Level of Service Inventory–Revised for Drug-Involved Offenders

Christopher E. Kelly and Wayne N. Welsh

Source: Kelly, C. E., & Welsh, W. N. (2008). The predictive validity of the Level of Service Inventory–Revised for drug-involved offenders. Criminal Justice and Behavior, 35(7), 819–831.

Questions for Thought

1. According to this study, what are the strongest predictors of reincarceration?

2. What are the specific limitations of this study?

3. Since the LSI-R had already been validated for a variety of different populations, why was it important

to validate this risk-assessment instrument for drug-involved offenders?

4. Why is it important to administer the LSI-R at multiple intervals? What are some of the potential problems associated with administering the LSI-R only at one point in time?


Do not

copy

, pos

t, or d

istrib

ute

Documents

3 RISK PREDICTION ASSESSMENT AND distribute...Chapter 3 • Assessment and Risk Prediction 53conviction and sentencing, the probation officer will complete the PSI, which will include